1. Field of the Invention
The present invention relates to a design aid method and an design aid apparatus for facilitating design of integrated circuits, and particularly to an LSI design aid method and an apparatus thereof for facilitating and shortening such design work, which can realize data input onto the same editorial image plane by means of both diagrams and texts while making these two kinds of data be constructionally related to one another.
2. Description of the Prior Art
In conventional methods for designing integrated circuits by aid of computers, data input by means of diagrams and text description based on a language in accordance with a predetermined grammar is generally used.
In such a design method, since each integrated circuit is designed with aid of computers, the method of data input depends on processing ability of an input tool to be used therein. However, since current input tools for data input by means of diagrams, such as graphic editors, have insufficient data input ability, in the initial step of designing an integrated circuit, inputting data by means of text description has been exclusively used.
Using text description methods, it is possible to express a wide variety of circuit states from circuit structures to abstract circuit states and algorithms under various conditions. Text description methods also allow a series of control systems generally having a complex structure for controlling the data path of the design circuit to be clearly expressed, so that it becomes unnecessary to design a circuit for the control system separately.
However, in the text description input method, since only character strings are displayed on the monitor, it is difficult for a designer to understand the contents at sight. Therefore, it is difficult to grasp the whole structure and operation of a circuit to be designed, so that it becomes also difficult to understand the data flow in the data path. Thus, the circuit structure and operation can not be satisfactorily visually verified by this method.
Moreover, when the text description input method is used, it is necessary to select one language suitable for each design specification from a plurality of languages and to master the grammar of the language. Therefore, it takes much time to make a design using this method.
Accordingly, methods of inputting data by means of a diagram input tool have been considered recently.
According to such a diagram input method, since a circuit to be designed is displayed on the monitor as a diagram, it is easy for a designer to understand the structure and state of the circuit. Therefore, it becomes possible to grasp the whole structure of the circuit, and also possible to satisfactorily carry out visual verification of the circuit operation.
However, as mentioned above, methods of inputting data by means of diagram input tool are still unsatisfactory in that the input work thereof is complex as compared with the text description input method. Moreover, as compared with the text description input method, it is difficult for such a diagram input method to express highly abstract concepts concerning circuit states and algorithms.
In the diagram input method, design data are divided into data sheets corresponding to each design process. The division of a design circuit into data sheets is carried out based on the physical structure thereof. However, it is impossible to divide based on the concepts of the circuit states and signals flowing therein.
In the diagram input method, the above-mentioned control system must be provided in a circuit to be designed and the data path system and the control system must be expressed together in one circuit diagram, so that the discrimination therebetween is very difficult.
An apparatus which uses both the diagram input method and the text description input method has been recently proposed. However, in such an apparatus, even though it becomes possible to carry out data input by means of both of these methods, it is not still impossible to carry out data edition on the same editorial image plane. Therefore, it is necessary to provide separate design editors for these two method. Because the data inputted by the diagram input method and those inputted by the text description input method have no relation with one another, it is very difficult to facilitate the design work.
As stated above, in the conventional LSI design aid apparatus which uses either the diagram input method or the text description input method, it is very difficult to facilitate the design work.
On the other hand, in the apparatus which uses both of the diagram and the text description input method, the design editors and data construction each corresponding to the respective methods must be provided independently. Therefore, it is impossible to grasp the mutual relation between data inputted by the respective methods.
When text descriptions and diagrams are inputted on the same image plane, diagrams are inputted onto a description input area, or a description input area is designated in a diagram input area, so as to input description data into the designated area. In the former case, since character strings are all dealt with as data in the description input area, positions where characters are to be inputted are restricted by the character pitch specified in the description area, so that it is not permissible to input characters to any given position. In the latter case, each description input area is designated in each diagram input area at any given position with any given size, so as to input description data thereto. In such a case, the designated area is dealt with as one diagram on a corresponding diagram input area, otherwise it is designated as a completely new window.
When dealt with as one diagram on a diagram input area, the description input area is fixed, and it is impossible to input data on the character number per one line or line number which exceeds the designated area. To make this possible, it is necessary to provide some process for area extension. Moreover, in such a case, if the diagram is enlarged or reduced in size, the size of inputted characters must be changed in accordance with the change of diagram. Thus, when enlarged excessively, it becomes difficult to understand all the characters in sight, and when reduced extremely, it becomes difficult to read the characters.
On the other hand, when the above-mentioned designated area is dealt with as a completely new window, each description input area is completely independent of diagram input areas. Thus, there is no positional relation between these two kinds of areas. Therefore, it is impossible to arrange diagrams having some relation to some description in the vicinity of the description.
In the text description input method, each description should be expressed with a language in accordance with a predetermined grammar. Therefore, in some conventional apparatus, when the input content corresponding to each description is stored in a data base, a command such as COMPILE/LINK is executed after all descriptions are inputted. In such an apparatus, since there is a considerable time interval between the description input and error check by means of a compiler or output of the error message, it is necessary as the designer to memorize the meaning of description at a position to which error occurrence is indicated, or to make a check again on the specification thereof. Moreover, when a part of description is corrected, all the descriptions must be compiled again, thus, the working efficiency for debugging deteriorates.
Namely, in the text description method, since it is possible to execute the command such as COMPILE/LINK only after all descriptions are inputted and with a minute correction the COMPILE/LINK command have to be executed again, the debugging efficiency is very bad.